def parse_cdhit_output_100p(unclustered_fasta, cdhit_output):
final_cluster_mappings = dict()
seqdict_100p = dict()
for rec in SeqIO.parse(cdhit_output, 'fasta'):
fwd = str(rec.seq)
rev = misc.revcomp(fwd)
seq_key = tuple(sorted([fwd, rev]))
seqdict_100p[seq_key] = rec
for rec in SeqIO.parse(unclustered_fasta, 'fasta'):
fwd = str(rec.seq)
rev = misc.revcomp(fwd)
seq_key = tuple(sorted([fwd, rev]))
cluster = seqdict_100p[seq_key].id
final_cluster_mappings[rec.id] = cluster
return final_cluster_mappings
def get_inferred_sequences(pairs, genome_dict, add_softclipped_bases=False):
inferred_sequences = []
for read1, read2 in pairs:
if read1.query_name.count('_') == 2:
context_width = int(read1.query_name.split('_')[-2])
name = read1.reference_name + ':' + str(
read1.reference_start +
context_width) + '-' + str(read2.reference_end - context_width)
inferred_sequence = genome_dict[read1.reference_name][
read1.reference_start:read2.reference_end]
if add_softclipped_bases:
inferred_sequence = sctools.left_softclipped_sequence_strict(
read1
) + inferred_sequence + sctools.right_softclipped_sequence_strict(
read2)
inferred_sequence = inferred_sequence[context_width:-context_width]
if read1.query_name.split('_')[-1] == '2':
inferred_sequence = misc.revcomp(inferred_sequence)
contig_edge = False
if sctools.is_left_softclipped_strict(read1) and \
sctools.left_softclipped_position(read1) < 0:
contig_edge = True
elif sctools.is_right_softclipped_strict(read2) and \
sctools.right_softclipped_position(read2) >= len(genome_dict[read2.reference_name]):
contig_edge = True
else:
name = read1.reference_name + ':' + str(
read1.reference_start) + '-' + str(read2.reference_end)
inferred_sequence = genome_dict[read1.reference_name][
read1.reference_start:read2.reference_end]
if add_softclipped_bases:
inferred_sequence = sctools.left_softclipped_sequence_strict(
read1
) + inferred_sequence + sctools.right_softclipped_sequence_strict(
read2)
if read1.query_name.split('_')[-1] == '2':
inferred_sequence = misc.revcomp(inferred_sequence)
contig_edge = False
if sctools.is_left_softclipped_strict(read1) and \
sctools.left_softclipped_position(read1) < 0:
contig_edge = True
elif sctools.is_right_softclipped_strict(read2) and \
sctools.right_softclipped_position(read2) >= len(genome_dict[read2.reference_name]):
contig_edge = True
inferred_sequences.append(
(name, len(inferred_sequence), contig_edge, inferred_sequence))
return inferred_sequences
def get_inferred_sequence(self, forward_read, reverse_read, is_reverse):
contig = forward_read.reference_name
start = forward_read.reference_start
end = reverse_read.reference_end
inferred_sequence = ''.join(self.genome_dict[contig][start:end])
inferred_sequence = sctools.left_softclipped_sequence_strict(forward_read) + \
inferred_sequence + \
sctools.right_softclipped_sequence_strict(reverse_read)
inferred_sequence = inferred_sequence[self.context_width:-self.context_width]
if is_reverse:
inferred_sequence = misc.revcomp(inferred_sequence)
contig_edge = False
if sctools.is_left_softclipped_strict(forward_read) and \
sctools.left_softclipped_position(forward_read) < 0:
contig_edge = True
elif sctools.is_right_softclipped_strict(reverse_read) and \
sctools.right_softclipped_position(reverse_read) >= len(self.genome_dict[contig]):
contig_edge = True
return inferred_sequence, contig_edge
def get_inferred_sequence(self, forward_read, reverse_read, is_reverse):
contig, start, end = forward_read.reference_name, forward_read.reference_start, reverse_read.reference_end
inferred_sequence = ''.join(self.genome_dict[contig][start:end])
inferred_sequence = sctools.left_softclipped_sequence_strict(forward_read) + \
inferred_sequence + \
sctools.right_softclipped_sequence_strict(reverse_read)
if is_reverse:
inferred_sequence = misc.revcomp(inferred_sequence)
return inferred_sequence
def add_sequence_to_secondary_alignment(sam_file_in, sam_file_out):
outfile = open(sam_file_out, 'w')
infile = pysam.AlignmentFile(sam_file_in, 'r')
outfile = pysam.AlignmentFile(sam_file_out, "w", template=infile)
current_seq = None
current_seq_reverse = None
for read in infile:
if read.query_sequence is not None:
current_seq = read.query_sequence
current_seq_reverse = read.is_reverse
else:
if current_seq_reverse == read.is_reverse:
read.query_sequence = current_seq
else:
read.query_sequence = misc.revcomp(current_seq)
outfile.write(read)
outfile.close()
def get_seq_lengths(clusters,
seqs,
header=[
'cluster', 'num_unique_seqs', 'mean_length',
'min_length', 'max_length'
]):
seq_lengths1 = defaultdict(set)
for cluster, seq in zip(clusters, seqs):
seq_lengths1[cluster].add(seq)
seq_lengths2 = defaultdict(
lambda: {
'unique_seqs': set(),
'num_unique_seqs': 0,
'mean_length': 0,
'min_length': 0,
'max_length': 0
})
for cluster in seq_lengths1:
for seq in seq_lengths1[cluster]:
seq_lengths2[cluster]['unique_seqs'].add(
tuple(sorted([seq, misc.revcomp(seq)])))
for cluster in seq_lengths2:
all_seq_lengths = list(
map(lambda x: len(x[0]), seq_lengths2[cluster]['unique_seqs']))
seq_lengths2[cluster]['num_unique_seqs'] = len(all_seq_lengths)
seq_lengths2[cluster]['mean_length'] = np.mean(all_seq_lengths)
seq_lengths2[cluster]['min_length'] = np.min(all_seq_lengths)
seq_lengths2[cluster]['max_length'] = np.max(all_seq_lengths)
seq_lengths2 = [
tuple([cluster] + [seq_lengths2[cluster][lab] for lab in header[1:]])
for cluster in seq_lengths2
]
seq_lengths2 = pd.DataFrame(seq_lengths2, columns=header)
return seq_lengths2
def cluster_100p(infile, outfile):
seqdict1 = dict()
for rec in SeqIO.parse(infile, 'fasta'):
seq = str(rec.seq)
if seq not in seqdict1:
seqdict1[seq] = rec
seqdict2 = dict()
for seq in seqdict1:
fwd = seq
rev = misc.revcomp(seq)
seq_key = tuple(sorted([fwd, rev]))
if seq_key not in seqdict2:
seqdict2[seq_key] = seqdict1[seq]
with open(outfile, "w") as handle:
SeqIO.write(seqdict2.values(), handle, "fasta")
def write_termini_to_unpaired_fasta(pairs, fasta_prefix):
fiveprime_outseqs = []
threeprime_outseqs = []
for p in pairs:
name5p = p['pair_id']+'_1'
name3p = p['pair_id']+'_2'
seq5p = p['seq_5p']
seq3p = p['seq_3p']
record5p = SeqRecord(Seq(seq5p, IUPAC.IUPACAmbiguousDNA),
id=str(name5p), description=str(name5p)+'_5p')
record3p = SeqRecord(Seq(revcomp(seq3p), IUPAC.IUPACAmbiguousDNA),
id=str(name3p), description=str(name3p)+'_3p')
fiveprime_outseqs.append(record5p)
threeprime_outseqs.append(record3p)
with open(fasta_prefix+'.fasta', 'w') as output_handle:
SeqIO.write(fiveprime_outseqs, output_handle, 'fasta')
with open(fasta_prefix + '.fasta', 'a') as output_handle:
SeqIO.write(threeprime_outseqs, output_handle, 'fasta')
def get_inferred_sequences(pairs, genome_dict, add_softclipped_bases=False):
inferred_sequences = []
for read1, read2 in pairs:
name = read1.reference_name + ':' + str(
read1.reference_start) + '-' + str(read2.reference_end)
inferred_sequence = genome_dict[
read1.reference_name][read1.reference_start:read2.reference_end]
if add_softclipped_bases:
inferred_sequence = sctools.left_softclipped_sequence_strict(
read1
) + inferred_sequence + sctools.right_softclipped_sequence_strict(
read2)
if read1.is_read2:
inferred_sequence = misc.revcomp(inferred_sequence)
inferred_sequences.append(
(name, len(inferred_sequence), inferred_sequence))
return inferred_sequences